The Translator Reads Harder Than the Critic

Three reviewers praised the melancholic repetition in Samira Toumi's "Le Sel des Absents," and none of them noticed the repetition was a countdown, because none of them had to translate it.

The phrase "elle comptait les jours," she counted the days, appears in the novel exactly nine times, always attached to the same character, always in a slightly shorter sentence than the time before. Reviewers read this as atmosphere, a grieving woman's habit of mind rendered through repetition, and wrote sentences about rhythm and mourning that were not wrong so much as incomplete. When I sat down to translate the book, I had to decide, for each of the nine instances, exactly how much shorter the sentence should get, and that forced me to count them properly for the first time. Nine repetitions, shrinking at a steady rate, land the reader on the shortest version of the sentence in the exact chapter the missing man returns. It was never atmosphere. It was a clock, built into the prose, and no one reviewing the book in French had needed to notice it to write a competent review.

The adversarial read

A critic reads a novel to evaluate it, which means reading for effect: did this land, did that move me, is the prose doing its job. A translator reads a novel to rebuild it, sentence by sentence, in a system that does not share the original's grammar, its cognates, or its rhythm, which means reading for dependency: what does this sentence need from the sentence before it, what does it owe to the sentence after it, what would break if I changed its length by two words. I call this the adversarial read, because you are effectively trying to find every place the text will fight you before you commit to a version, and a text under that kind of pressure reveals structure that a sympathetic, evaluative reading has no reason to go looking for.

This is not a claim that translators are more attentive readers than critics in general. It is a claim about what the job forces you to notice. A critic can call a repeated phrase melancholic and move on, because the review does not require them to reproduce the exact shrinking rate of nine sentences across four hundred pages. A translator has no such option. Reproducing the pattern requires first extracting the pattern, counting it, timing it against the chapters, and that extraction is where Toumi's actual design surfaced, invisible to a reading that only had to feel the effect rather than rebuild the mechanism producing it.

What this changed about how I translated the ending

Once I understood the nine instances as a countdown rather than a mood, I stopped treating the final, shortest instance as a lyrical flourish and started treating it as the payoff of a structure the whole book had been building toward silently. My English version keeps the count at nine, keeps the shrinking rate proportional even though English syllables do not compress the same way French ones do, and lands the ninth instance in the same chapter, three words long, where the French version lands it at four. A translation that had treated the repetition as pure atmosphere, as every review up to that point implicitly had, would have felt free to add or drop an instance for the sake of English rhythm, and would have broken a clock nobody reviewing the book had told me was there to break.

The obvious pushback

Critics might object that this makes translators sound like they are discovering secret architecture in every repeated phrase, when sometimes a writer repeats an image simply because it pleases them, with no countdown, no clock, no design beyond the pleasure of the sound recurring. That pushback is correct often enough that I do not trust every pattern I find. My check is whether the pattern is doing measurable work: does the count matter, does the timing land on something structurally significant, or is it just recurrence for its own sake. Toumi's nine passed that test because the ninth lands precisely on the man's return. Not every repetition I have translated has passed it, and treating an ordinary stylistic habit as buried machinery is its own kind of overreading, one a good translator has to guard against as carefully as the critic's underreading.